Application of machine learning in combination with mechanistic modeling to predict plasma exposure of small molecules DOI Creative Commons
Panteleimon D. Mavroudis, Donato Teutonico, Alexandra Abós

и другие.

Frontiers in Systems Biology, Год журнала: 2023, Номер 3

Опубликована: Июнь 20, 2023

Prediction of a new molecule’s exposure in plasma is critical first step toward understanding its efficacy/toxicity profile and concluding whether it possible first-in-class, best-in-class candidate. For this prediction, traditional pharmacometrics use variety scaling methods that are heavily based on pre-clinical pharmacokinetic (PK) data. We here propose novel framework which preclinical prediction performed by applying machine learning (ML) tandem with mechanism-based modeling. In our proposed method, relationship initially established between molecular structure physicochemical (PC)/PK properties using ML, then the ML-driven PC/PK parameters used as input to mechanistic models ultimately predict candidates. To understand feasibility framework, we evaluated number (1-compartment, physiologically (PBPK)), PBPK distribution (Berezhkovskiy, PK-Sim standard, Poulin Theil, Rodgers Rowland, Schmidt), parameterizations (using vivo , or vitro clearance). most scenarios tested, results demonstrate PK profiles can be adequately predicted framework. Our analysis further indicates some limitations when liver microsomal intrinsic clearance (CLint) only pathway underscores necessity investigating variability emanating from different providing predictions. The suggested approach aims at earlier drug development process so decisions molecule screening, chemistry design, dose selection made early possible.

Язык: Английский

Artificial Intelligence in Pharmaceutical Technology and Drug Delivery Design DOI Creative Commons
Lalitkumar K. Vora, Amol D. Gholap, Keshava Jetha

и другие.

Pharmaceutics, Год журнала: 2023, Номер 15(7), С. 1916 - 1916

Опубликована: Июль 10, 2023

Artificial intelligence (AI) has emerged as a powerful tool that harnesses anthropomorphic knowledge and provides expedited solutions to complex challenges. Remarkable advancements in AI technology machine learning present transformative opportunity the drug discovery, formulation, testing of pharmaceutical dosage forms. By utilizing algorithms analyze extensive biological data, including genomics proteomics, researchers can identify disease-associated targets predict their interactions with potential candidates. This enables more efficient targeted approach thereby increasing likelihood successful approvals. Furthermore, contribute reducing development costs by optimizing research processes. Machine assist experimental design pharmacokinetics toxicity capability prioritization optimization lead compounds, need for costly animal testing. Personalized medicine approaches be facilitated through real-world patient leading effective treatment outcomes improved adherence. comprehensive review explores wide-ranging applications delivery form designs, process optimization, testing, pharmacokinetics/pharmacodynamics (PK/PD) studies. an overview various AI-based utilized technology, highlighting benefits drawbacks. Nevertheless, continued investment exploration industry offer exciting prospects enhancing processes care.

Язык: Английский

Процитировано

456
Machine Learning and Artificial Intelligence in Toxicological Sciences DOI Open Access
Zhoumeng Lin, Wei-Chun Chou

Toxicological Sciences, Год журнала: 2022, Номер 189(1), С. 7 - 19

Опубликована: Июль 21, 2022

Abstract Machine learning and artificial intelligence approaches have revolutionized multiple disciplines, including toxicology. This review summarizes representative recent applications of machine in different areas toxicology, physiologically based pharmacokinetic (PBPK) modeling, quantitative structure-activity relationship modeling for toxicity prediction, adverse outcome pathway analysis, high-throughput screening, toxicogenomics, big data, toxicological databases. By leveraging approaches, now it is possible to develop PBPK models hundreds chemicals efficiently, create silico predict a large number with similar accuracies compared vivo animal experiments, analyze amount types data (toxicogenomics, high-content image etc.) generate new insights into mechanisms rapidly, which was impossible by manual the past. To continue advancing field sciences, several challenges should be considered: (1) not all are equally useful particular type toxicology thus important test methods determine optimal approach; (2) current prediction mainly on bioactivity classification (yes/no), so additional studies needed intensity effect or dose-response relationship; (3) as more become available, crucial perform rigorous quality check infrastructure store, share, analyze, evaluate, manage data; (4) convert user-friendly interfaces facilitate their both computational bench scientists.

Язык: Английский

Процитировано

79
Revolutionizing drug formulation development: The increasing impact of machine learning DOI
Zeqing Bao,

Jack Bufton,

Riley J. Hickman

и другие.

Advanced Drug Delivery Reviews, Год журнала: 2023, Номер 202, С. 115108 - 115108

Опубликована: Сен. 27, 2023

Язык: Английский

Процитировано

56
Advances in artificial intelligence for drug delivery and development: A comprehensive review DOI
Amol D. Gholap, Md Jasim Uddin, Md. Faiyazuddin

и другие.

Computers in Biology and Medicine, Год журнала: 2024, Номер 178, С. 108702 - 108702

Опубликована: Июнь 7, 2024

Язык: Английский

Процитировано

51
The Role of “Physiologically Based Pharmacokinetic Model (PBPK)” New Approach Methodology (NAM) in Pharmaceuticals and Environmental Chemical Risk Assessment DOI Open Access
Deepika Deepika, Vikas Kumar

International Journal of Environmental Research and Public Health, Год журнала: 2023, Номер 20(4), С. 3473 - 3473

Опубликована: Фев. 16, 2023

Physiologically Based Pharmacokinetic (PBPK) models are mechanistic tools generally employed in the pharmaceutical industry and environmental health risk assessment. These recognized by regulatory authorities for predicting organ concentration–time profiles, pharmacokinetics daily intake dose of xenobiotics. The extension PBPK to capture sensitive populations such as pediatric, geriatric, pregnant females, fetus, etc., diseased those with renal impairment, liver cirrhosis, is a must. However, current modelling practices existing not mature enough confidently predict these populations. A multidisciplinary collaboration between clinicians, experimental modeler scientist vital improve physiology calculation biochemical parameters integrating knowledge refining models. Specific covering compartments cerebrospinal fluid hippocampus required gain understanding about xenobiotic disposition sub-parts. model assists building quantitative adverse outcome pathways (qAOPs) several endpoints developmental neurotoxicity (DNT), hepatotoxicity cardiotoxicity. Machine learning algorithms can physicochemical develop silico where data unavailable. Integrating machine carries potential revolutionize field drug discovery development risk. Overall, this review tried summarize recent developments in-silico models, qAOPs use improving along perspective. This act guide toxicologists who wish build their careers kinetic modeling.

Язык: Английский

Процитировано

50
An artificial intelligence-assisted physiologically-based pharmacokinetic model to predict nanoparticle delivery to tumors in mice DOI Creative Commons
Wei-Chun Chou, Qiran Chen, Long Yuan

и другие.

Journal of Controlled Release, Год журнала: 2023, Номер 361, С. 53 - 63

Опубликована: Июль 31, 2023

The critical barrier for clinical translation of cancer nanomedicine stems from the inefficient delivery nanoparticles (NPs) to target solid tumors. Rapid growth computational power, new machine learning and artificial intelligence (AI) approaches provide tools address this challenge. In study, we established an AI-assisted physiologically based pharmacokinetic (PBPK) model by integrating AI-based quantitative structure-activity relationship (QSAR) with a PBPK simulate tumor-targeted efficiency (DE) biodistribution various NPs. QSAR was developed using deep neural network algorithms that were trained datasets published "Nano-Tumor Database" predict input parameters model. optimized NP cellular uptake kinetic used maximum (DEmax) DE at 24 (DE24) 168 h (DE168) different NPs in tumor after intravenous injection achieved determination coefficient R2 = 0.83 [root mean squared error (RMSE) 3.01] DE24, 0.56 (RMSE 2.27) DE168, 0.82 3.51) DEmax. AI-PBPK predictions correlated well available experimentally-measured profiles tumors (R2 ≥ 0.70 133 out 288 datasets). This provides efficient screening tool rapidly on its physicochemical properties without relying animal training dataset.

Язык: Английский

Процитировано

48
Machine intelligence-accelerated discovery of all-natural plastic substitutes DOI Creative Commons
Tianle Chen, Zhenqian Pang, Shuaiming He

и другие.

Nature Nanotechnology, Год журнала: 2024, Номер 19(6), С. 782 - 791

Опубликована: Март 18, 2024

Abstract One possible solution against the accumulation of petrochemical plastics in natural environments is to develop biodegradable plastic substitutes using components. However, discovering all-natural alternatives that meet specific properties, such as optical transparency, fire retardancy and mechanical resilience, which have made successful, remains challenging. Current approaches still rely on iterative optimization experiments. Here we show an integrated workflow combines robotics machine learning accelerate discovery with programmable optical, thermal properties. First, automated pipetting robot commanded prepare 286 nanocomposite films various properties train a support-vector classifier. Next, through 14 active loops data augmentation, 135 nanocomposites are fabricated stagewise, establishing artificial neural network prediction model. We demonstrate model can conduct two-way design task: (1) predicting physicochemical from its composition (2) automating inverse fulfils user-specific requirements. By harnessing model’s capabilities, several substitutes, could replace non-biodegradable counterparts exhibiting analogous Our methodology integrates robot-assisted experiments, intelligence simulation tools eco-friendly starting building blocks taken generally-recognized-as-safe database.

Язык: Английский

Процитировано

37
Revolutionizing Drug Discovery: A Comprehensive Review of AI Applications DOI Creative Commons

Rushikesh Dhudum,

Ankit Ganeshpurkar, Atmaram Pawar

и другие.

Drugs and Drug Candidates, Год журнала: 2024, Номер 3(1), С. 148 - 171

Опубликована: Фев. 13, 2024

The drug discovery and development process is very lengthy, highly expensive, extremely complex in nature. Considering the time cost constraints associated with conventional discovery, new methods must be found to enhance declining efficiency of traditional approaches. Artificial intelligence (AI) has emerged as a powerful tool that harnesses anthropomorphic knowledge provides expedited solutions challenges. Advancements AI machine learning (ML) techniques have revolutionized their applications development. This review illuminates profound influence on diverse aspects encompassing drug-target identification, molecular properties, compound analysis, development, quality assurance, toxicity assessment. ML algorithms play an important role testing systems can predict such pharmacokinetics candidates. not only strengthens theoretical foundation this technology, but also explores myriad challenges promising prospects combination offers strategy overcome complexities pharmaceutical industry.

Язык: Английский

Процитировано

31
Advancing Drug Safety in Drug Development: Bridging Computational Predictions for Enhanced Toxicity Prediction DOI Creative Commons
Ana M. B. Amorim, Luiz F. Piochi, Ana Teresa Gaspar

и другие.

Chemical Research in Toxicology, Год журнала: 2024, Номер 37(6), С. 827 - 849

Опубликована: Май 17, 2024

The attrition rate of drugs in clinical trials is generally quite high, with estimates suggesting that approximately 90% fail to make it through the process. identification unexpected toxicity issues during preclinical stages a significant factor contributing this high failure. These can have major impact on success drug and must be carefully considered throughout development late-stage rejections or withdrawals candidates significantly increase costs associated development, particularly when detected after market release. Understanding drug-biological target interactions essential for evaluating compound safety, as well predicting therapeutic effects potential off-target could lead toxicity. This will enable scientists predict assess safety profiles more accurately. Evaluation critical aspect biomolecules, proteins, play vital roles complex biological networks often serve targets various chemicals. Therefore, better understanding these crucial advancement development. computational methods protein–ligand emerging promising approach adheres 3Rs principles (replace, reduce, refine) has garnered attention recent years. In review, we present thorough examination latest breakthroughs prediction, highlighting significance drug-target binding affinity anticipating mitigating possible adverse effects. doing so, aim contribute effective secure drugs.

Язык: Английский

Процитировано

20
A Machine Learning Model to Estimate Toxicokinetic Half-Lives of Per- and Polyfluoro-Alkyl Substances (PFAS) in Multiple Species DOI Creative Commons
Daniel E. Dawson, Christopher Lau, Prachi Pradeep

и другие.

Toxics, Год журнала: 2023, Номер 11(2), С. 98 - 98

Опубликована: Янв. 20, 2023

Per- and polyfluoroalkyl substances (PFAS) are a diverse group of man-made chemicals that commonly found in body tissues. The toxicokinetics most PFAS currently uncharacterized, but long half-lives (t½) have been observed some cases. Knowledge chemical-specific t½ is necessary for exposure reconstruction extrapolation from toxicological studies. We used an ensemble machine learning method, random forest, to model the existing vivo measured across four species (human, monkey, rat, mouse) eleven PFAS. Mechanistically motivated descriptors were examined, including two types surrogates renal transporters: (1) physiological descriptors, kidney geometry, transporter expression (2) structural similarity defluorinated endogenous affinity. developed classification (Bin 1: <12 h; Bin 2: <1 week; 3: <2 months; 4: >2 months). had accuracy 86.1% contrast 32.2% y-randomized null model. A total 3890 compounds within domain model, was predicted using bin medians: 4.9 h, 2.2 days, 33 3.3 years. For human t½, 56% classified 4, 7% 3, 37% 2. This synthesizes limited available data allow tentative prioritization.

Язык: Английский

Процитировано

28